The long-term goals of this research are: 1) to further elucidate the relationships among variables which affect breast feeding behaviors; 2) to study relationships between breast feeding behaviors and selected maternal and infant outcomes; 3) to contribute to the body of knowledge specific to sub-groups of women who are less likely to choose to breast feed or are more likely to breast feed with limited success; 4) and to test the effectiveness and efficiency of interventions intended to promote and assist breast feeding among the most vulnerable sub- populations. The proposed study will address goals one, three, and four; work related to goal two will occur at a later stage in the program of research. Specific aims of the proposed longitudinal study follow. Test a hypothesized structural model of relationships between a set of predictor variables and selected maternal breast feeding behaviors within an experimental design framework. Subjects who are working mothers will be randomly assigned to experimental and control status with respect to an intervention program designed to ease the transition to working and breast feeding. Subjects who are not working outside the home constitute an additional control group. Objectives related to testing the model are: a) to determine whether the structural model fits for each group and is invariant across the groups; and b) if the model is invariant, to compare the adjusted means on the criterion variables across the groups. Develop a predictive model for breast feeding behaviors of working mothers. Predictor variables in the model will be those variables relevant to the working situation that were not included in the structural model (Aim 1). Initial data collection will occur during the postpartum hospitalization. Follow-up for all subjects will occur at 1, 3, 6, 9, adn 12 months. For those still breast feeding at 12 months, follow-up will continue at three month intervals, until weaning has occurred. Control and experimental interventions will be done by phone at 5 and 14 days postpartum and during a home visit at one month postpartum. Data analysis will include path analysis, multiple regression, and descriptive numerical and graphical techniques.